ENDANGERED SPECIES RESEARCH
Endang Species Res
Vol. Preprint, 2007
doi: 10.3354/esr00066
Published online December 7, 2007
OPEN
ACCESS
THEME SECTION
Post-nesting migrations of Galápagos
green turtles Chelonia mydas in relation to
oceanographic conditions: integrating satellite
telemetry with remotely sensed ocean data
Jeffrey A. Seminoff1,*, Patricia Zárate2, Michael Coyne3, David G. Foley4, 5,
Denise Parker1, Boyd N. Lyon6,†, Peter H. Dutton1
1
NOAA-National Marine Fisheries Service, Southwest Fisheries Science Center, 8604 La Jolla Shores Dr., La Jolla,
California 92037, USA
2
Charles Darwin Research Station, Puerto Ayora, Santa Cruz, Galápagos Islands, Ecuador
3
SEATURTLE.ORG, 1 Southampton Place, Durham, North Carolina 27705, USA
4
Joint Institute for Marine and Atmospheric Research, University of Hawai’i at Manoa, 1000 Pope Road, Honolulu,
Hawai’i 96822, USA
5
NOAA Fisheries, Environmental Research Division, Southwest Fisheries Science Center, 1352 Lighthouse Avenue,
Pacific Grove, California 93950-2097, USA
6
Department of Biology, University of Central Florida, PO Box 162368, Orlando, Florida 32816, USA
ABSTRACT: Post-nesting movements of 12 green turtles from the Galápagos Islands (Ecuador) were
tracked with satellite telemetry during the 2003 and 2005 nesting seasons. To illuminate potential
environmental influences on turtle movements we compared tracks with a variety of remotely sensed
oceanographic variables including sea surface temperature (SST), SST front probability, surface
height anomaly, surface current, and surface chlorophyll a concentration. Three distinct post-nesting
migratory strategies were observed, including oceanic migration to Central America (Type A1 movements, n = 3), residency within the Galápagos (Type A2 movements, n = 2), and movement into oceanic
waters southwest of the Galápagos (Type B movements, n = 7). Two turtles migrating to Central America reached neritic foraging areas in Nicaragua and Panama that were 1500 and 1542 km, respectively,
from their nesting sites, and one resident turtle established a foraging home range 75 km from its final
nesting site. Oceanic movements occurred in waters with a mean SST of 26.5°C and mean surface
chlorophyll a concentration of 0.18 mg m– 3, whereas neritic movements were in waters with a mean
SST of 24.3°C and mean surface chlorophyll a concentration of 0.47 mg m– 3. All turtles accessed SST
frontal zones at a greater rate than their availability, and at least 2 turtles conducted movements in the
oceanic zone that were indicative of foraging activity. This is the first report of migratory corridors for
Galápagos green turtles, confirming prior flipper tagging data that show that the Galápagos is a
source rookery for green turtles in coastal areas of Central America. The high proportion of green turtles departing the Galápagos (83%) indicates that marine fisheries bycatch and directed hunting on
this stock outside the Galápagos may impact this population more than previously believed, and
underscores the need for multi-national conservation efforts that combat these threats.
KEY WORDS: Black turtle · Chelonia agassizii · Cheloniidae · Chlorophyll a · Eastern Tropical Pacific
Ocean · Frontal zones · Migration · Sea surface temperature · Sea surface height anomaly
Resale or republication not permitted without written consent of the publisher
*Email: [email protected]
†
Deceased. Please see Dedication following Acknowledgements
© Inter-Research 2007 · www.int-res.com
2
Endang Species Res 3: preprint, 2007
INTRODUCTION
Green turtles Chelonia mydas are highly migratory
and undertake complex movements through geographically disparate habitats during their life cycle
(Musick & Limpus 1997, Plotkin 2003). The periodic
migration of hundreds to thousands of kilometers
between breeding and foraging areas by adults is a
prominent feature of their life history. The longdistance movements of green turtles were first resolved with flipper tagging (Carr et al. 1978, Alvarado
& Figueroa 1992) and more recently with genetic
analyses (Bass et al. 2006, Dethmers et al. 2006). The
use of satellite telemetry has expanded this understanding, revealing foraging area destinations of
migrating post-nesting turtles as well as providing
information on migratory routes, travel speeds, and
dive behavior (Cheng 2000, Godley et al. 2002, Hays et
al. 2002, Craig et al. 2004, Kennett et al. 2004, Troëng
et al. 2005, Blumenthal et al. 2006). Further, because
satellite telemetry yields spatio-temporal patterns of
turtle distribution, the resultant migratory tracks can
be integrated with remotely sensed data to elucidate
the underlying oceanographic conditions that may
influence sea turtle habitat use.
Understanding such factors is critical for improving
management and conservation initiatives for sea turtles. While basic environmental features such as
bathymetry, sea surface temperature (SST), and surface currents have been linked to sea turtle movements (Morreale et al. 1996, Hays et al. 2001, Luschi et
al. 2003, Pelletier et al. 2003, Etnoyer et al. 2006), the
mechanisms underlying these relationships are inadequately understood. Dynamic mesoscale processes
such as SST and chlorophyll fronts — areas of interface
between 2 dissimilar water masses — are known to
strongly affect water column primary and secondary
productivity (Olson et al. 1994, Palacios et al. 2006),
and their status as prey aggregation zones suggests
that they too may influence sea turtle movements. For
example, juvenile loggerhead turtles Caretta caretta in
the North Pacific have been shown to forage in the
Kuroshio Extension Bifurcation Region during periods
of high surface chlorophyll (Polovina et al. 2006); when
waters become vertically stratified and surface chlorophyll levels decrease, they move farther north to the
Transition Zone Chlorophyll Front region — a welldescribed biological hotspot and an important oceanic
foraging destination for a variety of marine taxa (e.g.
Polovina et al. 2000, 2001). Frontal zones have also
been described as important foraging zones for large
whales (Doniol-Valcroze et al. 2007) and commercially
targeted species such as tuna and swordfish (Fiedler &
Bernard 1987, Podesta et al. 1993). Knowledge of the
spatio-temporal persistence of these mesoscale fea-
tures can be highly informative for developing predictive models of sea turtle habitat use; however, there is
an obvious need for additional data from a variety of
marine species and ocean regions to determine the
efficacy of such modeling efforts.
The Eastern Tropical Pacific Ocean (ETP) is an area
with complex topography and substantial spatiotemporal variability in oceanographic characteristics
(Chavez et al. 1999, Fieldler 2002a). Oceanic waters
in this region contain several biological hotspots,
including the nutrient-enriched Galápagos plume
(Palacios 2002), the shoaling thermocline of the Costa
Rica Dome (Fiedler 2002b), and the Peruvian coastal
upwelling region (Strub et al. 1995, 1998). Such mesoscale features can be important areas of aggregation
for a variety of marine organisms (Worm et al. 2005,
Pennington et al. 2006), but their persistence is dramatically impacted by the El Niño Southern Oscillation
(ENSO). This 3 to 6 yr cycle within the coupled oceanatmosphere system of the tropical Pacific brings increased surface water temperatures and lower primary
productivity, both of which have profound biological
consequences (Chavez et al. 1999). This variability in
oceanographic conditions coupled with the presence of
numerous well-described biological hotspots provides
an ideal opportunity to examine the effects of ocean
climate on large, tractable marine species. As the ETP
is a region of high use by sea turtles, this taxon is a
prime candidate for commencing such studies. Satellite telemetry studies have described movements of
olive ridley turtles Lepidochelys olivacea and leatherback turtles Dermochelys coriacea in relation to
bathymetry in the area (Plotkin et al. 1995, Morreale et
al. 1996), but so far no studies have examined turtle
movements in relation to other oceanographic conditions in the ETP, although Saba et al. (2007) examined
leatherback nesting biology relative to oceanography
in the region.
Prominent in this area are the Galápagos Islands
(hereafter referred to as the Galápagos), a 19-island
archipelago located ~1000 km from coastal Ecuador
(Fig. 1). The Galápagos are among the most important
nesting areas for green turtles in the ETP, and turtles
are afforded protection by the 138 000 km2 Galápagos
Biological Reserve of Marine Resources, which regulates fishing pressures and provides marine habitat
and nesting beach protection (Heylings et al. 2002).
Flipper tagging programs (Green & Ortiz-Crespo 1982,
Green 1984, Zárate & Dutton 2002) and genetic data
(P. H. Dutton unpubl. data) have shown that green turtles nesting in the Galápagos move to foraging areas
both within the archipelago and along the continental
shelf of Central and South America. However,
although Galápagos green turtles evidently cross
oceanic areas to reach mainland coastal habitats,
Seminoff et al.: Post-nesting migrations of Chelonia mydas
nothing is known about their migratory corridors, nor
are there data on how these turtles are affected by
oceanographic conditions. Further, with fewer than 30
long distance (>1000 km) tag returns of the ~14 000
green turtles that have been tagged (Green 1984,
Zárate & Dutton 2002), there are sparse data on the relative importance of various coastal foraging sites for
this nesting stock.
Knowledge about the oceanic movements and foraging area destinations of Galápagos green turtles can
elucidate their susceptibility to human impacts such as
fisheries bycatch and targeted capture and can also
illuminate the areas of greatest need for conservation
efforts. In the present study, we use satellite telemetry
to track the post-nesting movements of green turtles
from the Galápagos and compare these movement
paths with a range of remotely sensed oceanic parameters. Our goals were to (1) quantify the proportion of
green turtles that depart the Galápagos after nesting
activity, (2) determine their migratory corridors and
foraging area destinations, and (3) determine which, if
any, oceanographic features influenced the movements of green turtles. Besides providing information
on green turtles per se, our results will also have general relevance for the understanding of oceanographic
influences on other sea turtle species in the ETP.
MATERIALS AND METHODS
Tracking data. Twelve adult
female green turtles ranging from
71.6 to 96.5 cm curved carapace
length (CCL, measured from nuchal
notch to posterior-most tip of carapace; mean = 83.0 ± 2.0 cm) were
fitted with satellite transmitters
(Platform Transmitter Terminals;
PTTs) during the 2003 (n = 4) and
2005 (n = 8) nesting seasons in the
Galápagos Archipelago, Ecuador
(Appendix 1). The turtles were
equipped with PTTs prior to returning to the sea at 3 different nesting
beaches: Las Salinas (Baltras Island;
n = 3), Barahona (Isabela Island; n =
3), and Quinta Playa (Isabela Island;
n = 6) (Fig. 1).
We used Telonics ST-18 (in 2003)
and ST-20 (in 2005) transmitters
(Mesa) that measured 6.0 × 12.3 ×
2.8 cm in size and 276 g in weight
(<1% of the approximate weight of
each turtle). For the ST-18 transmitters, duty cycles (hours during
3
which transmissions are turned on/off to maximize
battery life) were 12/48 for 3 of the PTTs and 24/24
for one of the PTTs. For the ST-20 transmitters, duty
cycles were 24/0 for Days 1 to 45, 12/25 for Days 46
to 70, and 6/25 for Day 71 to end of tag life. Battery
life was further prolonged by the presence of a saltwater-activated switch that prevented transmissions
during immersion. Transmitters were affixed to the
highest point of each turtle’s carapace using fiberglass cloth and polyester resin (Balazs et al. 1996)
after surfaces had been cleansed of grease and
debris. Each unit was oriented with the antenna to
the rear, with a roll of Kevlar® attached anterior to
the antenna to protect its base.
Turtle positions were determined with the Argos
system, which categorized each location message
received into one of 6 location classes (LC): 3, 2, 1,
0, A, and B (Appendix 1). Argos assigns accuracy estimations of <150 m for LC 3, 150 to 350 m for LC 2, 351
to 1000 m for LC 1, and >1000 m for LC 0. No accuracy
estimation is provided for LC A or LC B. However,
rather than limiting our track construction to the use
of LCs with optimal accuracy estimates (i.e. LC 3, 2,
and 1), we instead used a series of filters that excluded
biologically unreasonable results for travel speed
(> 5 km h–1) or indicated turning angles that did not
conform to a directional track line (<10°), as this
approach has been shown to maximize the utility of
Argos-derived positions for wildlife tracking (M. S.
Coyne et al. unpubl.).
Fig. 1. Economic Exclusive Zones (EEZs) in the Eastern Tropical Pacific with inset
map of the Galápagos Islands. Dashed lines indicate EEZ boundaries; Q: satellite
tag deployment locations — 1, Quina Playa (Isabela Island); 2, Barahona (Isabela
Island); 3, Las Salinas (Baltras Island)
4
Endang Species Res 3: preprint, 2007
Displacement (km)
(delineated by 13° N, 13° S, 102° W, and the coast of
The Argos locations were analyzed using Satellite
Central and South America). The FPI used herein is
Tracking and Analysis Tool (STAT) (Coyne & Godley
derived through a combination of gradient threshold
2005) and were mapped using MapTool and Generic
tests and along-front feature tracking algorithms (e.g.
Mapping Tools (Wessel & Smith 1991). The tracks from
each turtle were partitioned based on
the relationship between the time since
2000
deployment and distance from release
Type A2
Type A1
point (i.e. displacement plot; BlumenCM1
1500
CM9
thal et al. 2006; our Fig. 2). The curves
CM8
CM10
of these displacement plots showed
1000
CM12
distinct inflections that corresponded
with changes in travel speed as turtles
500
commenced and/or completed migratory movements. Post-nesting migra0
tions were considered to start on the
day coinciding with the beginning of a
2000
positive slope lasting > 25 d in the disType B-western
Type B-southern
placement curve and, for turtles moving
CM3
CM2
1500
to Central America, migration was
CM4
CM5
considered to have been completed on
CM6
CM11
1000
the day that the positive slope reached
CM7
asymptote. A migration straightness
500
index (MSI) was calculated for each
turtle’s migratory movements, based on
the ratio of straight line distance
0
100
100 0
0
25
50
75
25
50
75
between first and last oceanic Argos
locations to the total oceanic track
Deployment duration (d)
length (Luschi et al. 1998, Godley et al.
Fig. 2. Chelonia mydas. Displacement plots for green turtles tagged in the
2002). We determined mean travel
Galápagos Islands
–1
speed (km d ) in each zone using the
travel speeds of each successive track
Table 1. Satellite data extracted along the track of each turtle. These ocean
segment within the zone separated by
satellite data products were extracted from the OceanWatch Thematic Real≥12 h. Mean ± SE are given unless othtime Environmental Distributed Data System (THREDDS). MODIS: Moderate
erwise noted.
Resolution Imaging Spectrometer; GOES: Geostationary Operational EnvironSatellite data. We compared the
mental Spacecraft; POES: Polar-orbiting Operational Environmental Spacetracks of each turtle with information on
craft; AMSR: Advanced Microwave Scanning Radimeter; AVHRR: Advanced
Very-High Resolution Radiometer; GFO: Geosat Follow-On; OSCAR: Ocean
bathymetry, SST, SST front probability,
Surface Currents Analyses; SST: Sea surface temperature; FPI: Frontal Probageostrophic surface currents, sea surbility Index; SSHa: sea surface height anomaly. When >1 sensor is used, the
face height anomaly (SSHa), and surdata format is a blended product. In addition, bathymetric data were collected
face chlorophyll a (chl a) concentration.
via the General Bathymetric Chart of the Oceans (GEBCO), a global
Bathymetric values for the region were
topographic dataset with one minute (1’) spatial resolution
determined with the General Bathymetric Chart of the Oceans (GEBCO;
Parameter
Spacecraft Environmental
Spatial
Composite
www.ngdc.noaa.gov/mgg/gebco/; IOC,
sensor
grid
duration (d)
IHO, BODC 2003). SST data were
Surface chl a
Aqua
MODIS
0.05° × 0.05°
08
obtained from 4 sources (Table 1), and
SST
Aqua
MODIS
0.1°
×
0.1°
05
were blended with weighted averages
AMSR-E
0.1° × 0.1°
05
based on the nominal errors for each
GOES
Imager
0.1° × 0.1°
05
dataset. SST fronts were determined
POES
AVHRR
0.1° × 0.1°
05
using a Frontal Probability Index (FPI;
FPI
GOES
Imager
0.05° × 0.05°
014
Breaker et al. 2005, Castelao et al.
SSHa
Jason-1
GFO
0.25° × 0.25°
~10
2005), calculated as the fraction of days
Envisat
0.25° × 0.25°
~10
in a given 14 d period for which fronts
Ocean currents
OSCAR - real time 1.0° × 1.0°
010
are detected within each 0.05° × 0.05°
Bathymetry
GEBCO
1’
0–
(~5.5 × 5.5 km) cell in the study area
Seminoff et al.: Post-nesting migrations of Chelonia mydas
Canny 1986, Cayula & Cornillon 1995) applied to the
daily SST data obtained from the Imager (Table 1). The
advantage of this dataset over other SST data sources
is the high frequency of the images (48 d–1), which
greatly helps to mitigate the obscuring effects of clouds
(Castelao et al. 2005). Sea surface current (geostrophic
velocity vector) and SSHa data were obtained from the
Centre National d’études Spatiales Aviso/Altimetry
project (www.aviso.oceanobs.com; Ducet et al. 2000).
Satellite-derived chl a pigment concentration data
were collected by the Moderate Resolution Imaging
Spectrometer (MODIS) and obtained from NASA Goddard Space Flight Center, Ocean Color Project (http://
oceancolor.gsfc.nasa.gov). See Table 1 for a summary
of the spacecrafts, environmental sensors, and spatial
and temporal resolutions for each of the remotely
sensed oceanographic datasets.
Statistical analysis. Argos-derived migration paths
were compared to environmental conditions in 3
ways. To determine the water depth along each track
line, satellite tracks were overlaid on bathymetric
charts; locations in waters ≥200 m deep were considered ‘oceanic’, whereas those in water < 200 m were
considered neritic. To evaluate turtle movements in
relation to surface currents and SSHa, we constructed
animations in MapTool of the oceanic movements by
turtles tagged in 2005, with each frame representing
successive Argos turtle locations overlayed on the
spatio-temporally coincident oceanographic data (see
Figs. A1 to A7 in Appendix 2, available as Supplementary Material online at www.int-res.com/articles/
suppl/n003pp11_apps/). We determined the surface
chl a concentration, SSHa, SST, and SST frontal probability at each Argos-derived turtle location and also
calculated the mean for each oceanographic variable
within a box (0.2° longitude × 0.2° latitude × 5 to 10 d)
centered at the time and position of each satellite
transmission and taken to represent the available
habitat adjacent to each turtle location. The overall
mean values in neritic and oceanic zones for each
oceanographic variable were compared using 1-way
analysis of variance (ANOVA). We also used ANOVA
to compare FPI at Argos locations in the Galápagos
neritic, oceanic, and mainland neritic zones with the
mean FPI for each respective zone within the entire
study area. A non-parametric Kruskal-Wallis test was
used to confirm the results of the ANOVA for FPI, as
this index occurs in a non-continuous, non-normal
data distribution (Hollander & Wolfe 1999). Ocean
satellite data products used for ANOVA were
extracted from the OceanWatch Thematic Real-time
Environmental Distributed Data System (THREDDS)
hosted by the Environmental Research Division of the
NOAA Fisheries’ Southwest Fisheries Science Center
using the program Open-source Project for a Network
5
Data Access Protocol (OPeNDAP) and analyzed using
Matlab (Mathworks).
RESULTS
Three distinct post-nesting migratory strategies
were observed, the patterns of which we classify following the nomenclature of Godley et al. (2007, this
Theme Section [TS]). Three turtles conducted oceanic
migrations to access neritic foraging areas along the
Pacific coast of Central America (hereafter referred to
as Type A1 movements), 2 turtles remained in the
Galápagos (Type A2 movements), and 7 turtles moved
into oceanic waters southwest of the Galápagos (Type
B movements) (Table 2, Figs. 2 & 3). Among turtles
conducting Type B movements, there was a tendency
for their departures from the Galápagos to be largely
west- or south-bound (Type B-western and Type Bsouthern, respectively; Table 2, Figs. 2 & 3); however,
because we are unaware of the final destinations of
these turtles, all movements are classified as Type B
due to their mutual occurrence in oceanic waters.
There was a significant difference in body size among
groups (Kruskal-Wallis H = 31.93, p < 0.0001); turtles
engaging in Type A1 movements were the largest,
while those conducting Type A2 movements were the
smallest.
Type A1 movements
The 3 turtles that migrated to neritic habitats of
Central America (Turtles CM9, CM10, and CM12) conducted oceanic migrations that lasted from 31.3 to
53.7 d (Table 2). Turtle CM9, the largest tracked turtle
(96.5 cm CCL), moved northeast along the Cocos Ridge
for ~900 km and past the Nocoya Peninsula (Costa
Rica) en route to a neritic site just south of the Gulf of
Fonseca (Nicaragua). Turtle CM10 moved along a
direct route (MSI = 0.95) past Malpelo Island (Colombia) and into neritic waters of the Gulf of Panama.
Turtle CM12 traveled over the Cocos Ridge for
~1000 km, initially heading northwest away from the
Galápagos then turning northeast en route to the ridge
extending south of Panama, which it followed into the
Gulf of Chiriquí (Panama). The mean travel speeds
of turtles conducting Type A1 movements were significantly slower in neritic waters of Central America
(20.6 ± 6.0 km d–1) compared to oceanic waters (46.2 ±
5.5 km d–1) (t = 9.9, p = 0.03, Table 2). This decrease
was particularly evident for Turtles CM9 and CM10,
both of which we believe reached their foraging area
destinations based on the shorter daily travel distances
at the end of their tracks (Fig. 4).
6
Endang Species Res 3: preprint, 2007
Table 2. Post-nesting movement summary. Turtles are grouped according to post-nesting migratory strategy, and movements are
summarized as number of days (d), distance (km), and mean speed (km d–1) in each zone. Migration straightness index (MSI)
is provided for oceanic movements. CCL: curved carapace length (cm); (–): no data
Turtle
CCL
(cm)
Galápagos neritic
Duration Distance Speed
(km d–1)
Oceanic
Duration Distance Speed MSI
Mainland neritic
Overall
Duration Distance Speed Duration Distance
(km d–1)
Type A1
CM9
CM10
CM12
96.5
93.0
81.0
11.1
27.2
45.1
192.3
361.7
390.0
17.3
13.3
08.6
53.7
31.0
37.3
1894.2
1630.4
1901.8
35.2
52.6
50.9
0.79
0.95
0.70
5.3
7.5
7.8
049.9
171.9
233.1
09.4
22.8
29.7
70.2
65.7
90.3
2136.3
2164.0
2524.9
Type A2
CM1
CM8
74.7
71.6
47.9
69.0
1396.10
691.9
29.1
10.3
–
–
–
–
–
–
–
–
–
–
–
–
–
–
47.9
69.0
1396.1
0691.9
Type B-western
CM2
80.5
CM5
87.2
CM11
83.0
29.2
7.9
9.1
70
143.7
205.6
02.9
18.1
22.7
30.0
54.5
58.5
1231.7
1912.1
1534.8
41.0
35.1
26.2
0.95
0.50
0.87
–
–
–
–
–
–
–
–
–
59.1
62.5
67.6
1301.7
2055.8
1740.4
Type B-southern
CM3
89.0
CM4
86.5
CM6
82.6
CM7a
78.0
85.5
–
–
–
501.7
–
–
–
05.9
–
–
–
14.8
58.5
33.8
40.1
0397.4
1662.5
1404.5
1217.0
26.9
32.4
44.7
30.3
0.99
0.60
0.75
–
–
–
–
–
–
–
–
–
–
–
–
–
100.30
58.5
33.8
62.1
0899.1
1662.5
1404.5
1903.5
a
CM7 data based on partial track (Days 23 to 62); no MSI calculated
Type A2 movements
The 2 turtles remaining in the Galápagos (Turtles
CM1 and CM8) were tracked for 49.1 to 69.0 d and visited coastal areas of at least 5 different islands, but did
not nest again (Fig. 3). Turtle CM1 moved 1396.1 km
after departing the Barahona nesting beach, and traveled at a mean speed of 29.1 km d–1 during the tracking period (Table 2). Upon the final Argos transmission, Turtle CM1 was moving along the outer edge of
the Galápagos (water depth > 500 m) at a travel speed
> 45 km d–1 and had not yet associated with a neritic
foraging habitat. In contrast, Turtle CM8, the smallest
turtle tracked (71.6 cm CCL), moved to a neritic foraging area along the north coast of Floreana Island that
was 75 km from its nesting site on Isabela Island
(Fig. 3). The mean travel speed for Turtle CM8 during
its 49 d at this foraging area was 7.3 ± 1.2 km d–1, compared to a mean travel speed of 17.7 ± 2.9 km d–1 during the 19 d prior to its arrival at this site.
this route after first moving in waters south of the Galápagos for the first 49 d of its deployment. At their final
Argos-derived locations, these 3 turtles were at distances of 952 to 1331 km from the Galápagos and still
moving west at travel speeds of 25 to 50 km d–1 (Fig. 4).
The turtles conducting Type B-southern movements
(Turtles CM2, CM4, CM6, and CM7) were last tracked
at distances of 395 to 1247 km from the Galápagos
(Type B-southern in Fig. 3). However, the track for Turtle CM3 was too short to provide useful information,
while that of Turtle CM7 (the turtle tracked the farthest
south) was missing the initial 23 d of track due to transmitter malfunction, thus preventing a full understanding of this turtle’s oceanic movements. Nevertheless,
Turtles CM4, CM6, and CM7 were not always moving
away from the Galápagos while tracked and each
maintained their location in the oceanic zone for at
least 2 successive Argos transmissions on one or more
occasions. This is particularly evident for Turtle CM4,
which moved within a confined oceanographic area
centered on 8° S, 92° W for 15 d at the end of its tracking duration (Figs. 2 & 3).
Type B movements
The 7 turtles that departed the Galápagos to the
southwest did so within 0 to 85.5 d of tagging and were
tracked for 14.8 to 58.5 d in oceanic waters (Table 2).
Turtles CM2 and CM11 moved to the west along a consistent trajectory (MSI = 0.95 and 0.87, respectively;
Type B-western in Fig. 3). Turtle CM5 also followed
Oceanographic features
Integrating the oceanic movements of green turtles
with geostrophic surface currents indicates that their
movements oriented both with and against prevailing
currents. Geostrophic current animations for Type A1
Seminoff et al.: Post-nesting migrations of Chelonia mydas
7
Fig. 3. Chelonia mydas. Satellite-tracked post-nesting movements of green turtles nesting in the Galápagos. Circular labels
correspond to each of the 12 tracked turtles (i.e. 1 = CM1). Contour lines denote 2500 m water depth increments for Type A1
and Type B movements, and 500 m depth increments for Type A2 movements. Maps constructed with MapTool
movements depict the westward-flowing Northern
Equatorial Current (NEC) extending from the Galápagos to ~4° N, as well as the eastward-flowing Northern
Equatorial Countercurrent (NECC) south of the NEC
(see Figs. A1 to A3 in Appendix 2). During the initial
portions of northbound movements by Turtles CM9
(Fig. A1, Appendix 2) and CM10 (Fig. A2, Appendix 2),
both turtles swam against NECC counter-clockwise
rotating eddies (CM9 until 8° N and CM10 until 3.8° N)
and thereafter moved in concordance with northward
flow along the edge of counter-clockwise rotating
eddies. The initial movements of Turtle CM12 to the
northwest were in concordance with the NEC and at
~3° N this turtle turned to the northeast and traveled
largely in the same direction as surface currents, moving along the edges of a series of eddies as it progressed toward the coast of Central America (Fig. A3,
Appendix 2). The movements by Turtles CM9, CM10
and CM12 to the northwest upon arriving in neritic
waters may be reflective of compensatory navigation
8
Endang Species Res 3: preprint, 2007
Type A1
100
CM9
50
25
0
100
CM10
75
50
50
25
0
100
CM8
75
50
25
25
0
0
100
CM1
75
Travel speed (km d–1)
75
Travel speed (km d–1)
Type A2
100
0
20
40
60
80
100
Deployment duration (d)
CM12
75
50
Type B-southern
100
25
CM3
75
0
0
20
40
60
80
100
50
Deployment duration (d)
25
0
Type B-western
100
100
CM2
75
50
50
Travel speed (km d–1)
25
0
100
CM10
75
50
Travel speed (km d–1)
75
25
0
100
CM6
75
50
25
25
0
0
100
100
CM11
75
CM7
75
50
50
25
25
0
CM4
0
0
20
40
60
80
100
Deployment duration (d)
0
20
40
60
80
100
Deployment duration (d)
Fig. 4. Chelonia mydas. Swim speeds during satellite-tracked movement of green turtles in the Galápagos Islands
after being pushed to the east during migrations. If so,
this suggests that the east-bound NECC and its various
filaments were the most influential surface currents
north of the Galápagos.
Type B movements were also variably oriented to
prevailing surface currents. Turtles CM2 and CM11
(Fig. A4, Appendix 2) swam in concordance with
the westward-flowing Southern Equatorial Current
(SEC) for the entirety of their respective tracking
durations. Turtle CM5 also moved in parallel to the
SEC for the final 13.5 d of its track, after becoming
entrained in east-flowing surface waters and moving
southeast for the first 21 d after leaving the Galápagos
(Fig. A5, Appendix 2). At its southernmost location,
Turtle CM5 meandered for 11 d in a very small area
and its travel speed slowed to 34.3 ± 6.0 km d–1 com-
Seminoff et al.: Post-nesting migrations of Chelonia mydas
9
pared to 43.5 ± 6.1 km d–1 during earlier
movements and 50.8 ± 6.2 km d–1 during
later movements. The remaining Type B
movements were more variable in relation to surface currents and there was no
apparent major influence by the SEC or
other surface currents. Instead, Turtles
CM6 and CM7 (Figs. A6 & A7, respectively, Appendix 2) appeared to move
substantial distances against the prevailing currents.
As turtles moved between neritic and
oceanic zones, they encountered substantially different oceanographic conditions. A comparison of oceanographic
variables in waters adjacent to Argos
locations showed that these oceanic and
neritic areas had significantly different
values for SST, FPI, SSHa (p < 0.0001),
but were not different in terms of surface
chl a concentration, although we note
that this latter relationship was only marginally insignificant (p = 0.074; Table 3).
Likewise, the Argos-derived turtle locaFig. 5. Comparison of oceanographic variables at each Argos-derived green
tions in neritic and oceanic zones had
turtle location. (a,c) Surface chlorophyll a concentrations and (b,d) sea
significantly different values for SST,
surface temperature (SST) in the (a,b) oceanic and (c,d) neritic zones
SSHa, and chl a concentration (p < 0.05),
although there was no observed differrespect to the thermal environment, green turtles
ence for SST front probability (p = 0.76; Table 3). The
rarely occupied areas with SST ≤24.0°C, and mean
significant difference in mean surface chl a concentraSST along the oceanic portion of the turtles’ movetion at Argos locations in the neritic (0.47 ± 0.22 mg
ments was > 2°C warmer than that for their neritic
m– 3) and oceanic zones (0.18 ± 0.06 mg m– 3) despite the
consistency in mean concentration in available waters
movements (Fig. 5b,d). Selectivity for thermal condiof the neritic and oceanic zones (0.35 ± 0.007 and
tions was also indicated by the propensity of green
0.34 ± 0.001 mg m– 3, respectively) suggests that chl a
turtles to access SST frontal zones at rates greater than
concentrations may have influenced turtles differently
their availability; i.e. FPI at Argos locations was higher
in these 2 zones (Table 3). In support, the range of
than mean FPI in each respective zone (i.e. Galápagos
chl a concentrations experienced by turtles was
neritic, oceanic, Central America neritic) within the
much broader in the neritic zone (Fig. 5a,c). With
study area (Table 4).
Table 3. Comparisons of oceanographic conditions in neritic and oceanic zones for all Argos-derived turtle locations as well as the
means within a box (0.2° longitude × 0.2° latitude × 5 to 10 d) centered at the time and position of each satellite transmission and
taken to represent the available habitat (i.e. background) adjacent to each turtle location
Oceanographic
variable
Data set
n
Neritic
Mean
SE
n
Oceanic
Mean
SE
F-value Significance
(p)
Surface chl a
Argos locations
Background
146
5.6 × 104
0.47
0.35
0.014
0.007
194
2.7 × 106
0.18
0.339
0.013
0.001
237
3.2
< 0.0001
0.074
SST
Argos locations
Background
146
1391
24.27
24.23
0.146
0.07
194
6.4 × 104
26.51
26.31
0.128
0.010
133
924
< 0.0001
< 0.0001
FPI
Argos locations
Background
146
2642
0.051
0.0453
0.0078
0.0019
194
1.1 × 105
0.0475
0.0232
0.0067
0.0003
0.1
129
0.7556
< 0.0001
SSHa
Argos locations
Background
146
784
–0.0077
–0.0319
0.0038
0.0021
194
2.4 × 104
0.0051
–0.009
0.0033
0.0004
6.47
116
0.01
< 0.0001
10
Endang Species Res 3: preprint, 2007
the range of habitat types occupied during non-reproductive periods may contribute to their stability over geologic
time scales by buffering against catastrophic oceanographic events (e.g.
ENSO, red tide) that affect nesting frequency and abundance.
F
p
The observed variation in post-nesting migratory strategy raises intriguing
questions about the evolution of migra96.3 < 0.0001
tion in sea turtles. Although little is
known about how sea turtles choose
foraging sites, at least some green tur69.4 < 0.0001
tles are philopatric to specific foraging
habitats, returning to the same sites
45.3 < 0.0001
after nesting, year after year (Limpus et
al. 1992, Godley et al. 2002, Broderick
et al. 2007). If these are within the same
123
< 0.0001
ocean regions to which the turtles
recruit during hatchling and pelagic
juvenile dispersal, then explaining the
migration routes and preferred foraging areas would
benefit from focused examination of these dispersal
mechanisms. Surface currents have been shown to
play an important role in hatchling transport (Carr &
Meylan 1980, Bass et al. 2006), and the flow of 4 major
current systems (i.e. NEC, NECC, SEC, Peru Current)
near the Galápagos may facilitate long-distance dispersal in multiple directions. The numerous current
filaments and eddies generated by these convergences
may also result in dispersal over only a very short
distance from the nesting beach.
Upon recruitment to a foraging area, our results suggest that local conditions can profoundly affect the
growth and size-at-maturity of sea turtles. It is interesting to note that the 2 resident turtles (Type A2 movement
group) were the smallest turtles tracked during this
study. Although we urge caution due to the small sample
size, we suggest that their small body size may be due to
the high green turtle density in Galápagos foraging
habitats and the subsequent poor food availability
(Zárate 2007) from which the effects on growth were
perhaps enhanced as a result of foraging area philopatry.
Adverse density-dependent effects on green turtle
growth rate have been reported for green turtles in the
Bahamas (Bjorndal et al. 2000) and the fact that juvenile
growth rates in the Galápagos are among the slowest in
green turtles studied to date (≤0.5 cm yr–1; Green 1993)
suggests that density-dependent factors may be ongoing
in this region. Phenotypically linked dichotomies in
foraging areas, with larger turtles remaining in coastal
habitats and smaller turtles occupying oceanic foraging
zones, have been reported for loggerheads from the
Cape Verde Islands (Hawkes et al. 2006) and Japan
(Hatase et al. 2002). Consistent with our reasoning, the
Table 4. SST Frontal Probability Index (FPI) for Argos locations and background values for neritic and oceanic habitats in the Eastern Tropical Pacific.
FPI values were calculated as the fraction of days in a given 14 d period for
which fronts are detected within each 0.05° × 0.05° (~5.5 × 5.5 km) cell in the
study area (delineated by 13° N, 13° S, 102° W, and the coast of Central and
South America). All F values denote 95% significance
FPI location
Data type
N
All regions
Argos locations
199
0.0925 ± 0.007
Background
1.12 × 105 0.0237 ± 0.0003
Comparison
Galápagos
neritic
Argos locations
Background
Comparison
Oceanic
Argos locations
101
Background
1.09 × 105
Comparison
0.092 ± 0.01
0.024 ± 0.003
Mainland
neritic
Argos locations
Background
Comparison
0.138 ± 0.011
0.004 ± 0.005
85
2641
13
5400
Mean ± SD
0.087 ± 0.0131
0.0453 ± 0.0023
DISCUSSION
Migratory strategies
The mosaic of post-nesting migration strategies for
Galápagos green turtles, which included north-bound
migration to Central America, residency in the Galápagos, and dispersal to oceanic waters southwest of the
Galápagos, highlights the substantial flexibility in
migratory routes and foraging area destinations for this
nesting population. Variability of migratory strategies
and foraging habitat affinities within nesting populations is an emerging theme among satellite telemetry
studies of sea turtles (see review by Godley et al. 2007).
For green turtles, such flexibility has previously been
reported for nesting populations in Japan (Hatase et al.
2006) and the Mediterranean (Godley et al. 2003).
Among other hard-shelled turtles this trait has been
found in olive ridley turtles nesting in northern Australia, which migrate to coastal, continental shelf and
continental slope feeding habitats (Whiting et al.
2007). The occupation of diverse foraging habitats
after nesting is perhaps most common in loggerheads,
as dedicated oceanic and neritic foragers have been
documented among nesting populations in Japan
(Hatase et al. 2002) and the Cape Verde Islands
(Hawkes et al. 2006). Loggerheads nesting in North
Carolina, USA, also show foraging area flexibility, with
some females accessing temperate northern waters
during summer months and others remaining in more
southern, warm-water areas for the entirety of their
tracking periods (Hawkes et al. 2007). Although the
immense spatial scale of such movements may create
challenges for conserving these nesting populations,
Seminoff et al.: Post-nesting migrations of Chelonia mydas
smaller size of oceanic loggerheads from Japan is
believed to be the result of poorer food availability
(Hatase et al. 2002). Why smaller turtles would be present in Galápagos neritic versus oceanic waters as
reported by Hatase et al. (2002) is unresolved. However,
the smaller body size of resident turtles in our study,
coupled with the prevalence of Type B movements
(oceanic occupancy), suggests that oceanic foraging may
be more energetically beneficial than neritic foraging for
Galápagos green turtles. To further examine the energetic consequences of remaining in the Galápagos, it
would be interesting to compare the reproductive
periodicity of residents and migrants, since the rate of
nutrient uptake is a major determinant of the interbreeding interval (Bjorndal 1985).
Although prior flipper tagging efforts have demonstrated the movement of green turtles from the Galápagos to the coast of Central America (Green 1984),
this is the first study to reveal the specific migratory
corridors that are used. The Cocos Ridge and, to a
lesser extent, the ridge extending south of Panama
appear to be important features associated with green
turtle movements to Central America (Type A1 movement group). Although we do not know how these
bathymetric features are detected, they may promote
oceanographic conditions that are beneficial to migrating turtles. It is interesting that the Cocos Ridge also
demarks the primary migratory corridor for leatherback turtles departing nesting beaches in Costa Rica
(Morreale et al. 1996). However, it is also possible that
these ridges do not directly or indirectly influence
green turtle movements and are instead relevant only
because of their presence along the most direct routes
between the Galápagos and the preferred foraging
areas for each turtle. In the absence of surface current
influences on open-ocean migration, it is reasonable to
expect a nutrient-depleted post-nesting turtle to access
its foraging area by the most direct route possible in
order to minimize the energetic cost of migration.
Indeed, the Pacific coastal areas of Central America
are highly productive and host numerous foraging
hotspots for green turtles (Cornelius 1982, Amorocho &
Reina 2007).
Whereas green turtles moving to the north were en
route to neritic destinations along the coast of Central
America, the functions of green turtle movements to
the southwest of the Galápagos are less clear. We
acknowledge that some of these turtles may have
eventually returned to the Galápagos neritic, but
because all were still moving away from the Galápagos
at the last satellite transmission we believe this is
unlikely. Based on geostrophic current plots, it is
apparent that the westbound movements of turtles
departing the Galápagos in 2003 and 2005 were
assisted by the SEC, of which the surface layer from
11
~5° N to ~5° S is driven to the west by the prevailing
easterly trade winds. The maximum speed of the SEC
in this area reaches 25 cm s–1 (Kessler 2006) or 53 to
82% of the overall average speeds of Turtles CM2,
CM5, and CM11. The nearly identical west-bound
movements of these turtles may thus reflect surface
current-facilitated and bearing-mediated movements
towards known foraging areas outside the ETP. East
Pacific green turtles (based on mitochondrial DNA
data; Kuroyanagi et al. 1999) have been found in the
western Pacific, and turtles with western Pacific
mtDNA haplotypes are regularly sighted in the Galápagos (known locally as yellow turtles; Fritts 1981a, P.
H. Dutton unpubl. data), indicating that transPacific
movement of green turtles occurs in both directions. If
Turtles CM2, CM5, and CM11 continued along the
west –southwest trajectory, the nearest neritic habitats
they would encounter are in French Polynesia, still
~2500 km from where the 3 turtles last transmitted.
At the rate of travel measured when PTT transmissions ceased (26.2 to 41.0 km d–1), these turtles would
require an additional 70 to 95 d to reach this destination. This ~ 4000 km migration would represent one of
the longest post-nesting migrations known for green
turtles in terms of total distance and duration.
Linking satellite tracks with oceanoghraphy
Green turtles moving in oceanic zones as well as neritic zones in the Galápagos and Central America commonly associated with SST frontal zones. The affinity
to these frontal areas is underscored by the fact that
the ETP is a region of relatively low frontal activity
(Eden & Timmerman 2004). However, this result
should be viewed cautiously since there is a temporal
discontinuity between the background FPI calculation
(averaged over 14 d) and FPI at turtle locations (instantaneous for 1 d) in any one 0.05° × 0.05° cell. Nevertheless, although it is not certain that turtles associated
with fronts per se, they at least maintained affinity to
areas of more frequent front activity. The strategy used
by green turtles to associate with these areas is
unclear, although it is unlikely that they were able to
detect the fronts at distances of 10s to 100s of km and
actively navigate to them. Instead, we believe that turtles engaged with frontal zones opportunistically, and
remained in such areas only if prey resources were
adequate. Most fronts in this region result from current- or wind-induced upwellings, which increase biological productivity (Fiedler 2002a) and may result in
greater food abundance in surface waters for sea turtles and other higher order consumers. Thus, by
exploiting frontal areas, green turtles may decrease
prey search times and increase their feeding effi-
12
Endang Species Res 3: preprint, 2007
ciency. To further elucidate green turtle use of frontal
zones, it would be helpful to know more about their
diving profiles during association with these areas to
determine if their dives are shallower than when occupying waters away from thermal fronts.
Although SST frontal areas may constitute important
oceanic foraging habitats for green turtles, the absence
of east-bound movements toward the coast of South
America — an area of substantial frontal activity
(Fig. 6) — suggests that SST fronts may only be important if the associated temperatures are within an optimal range. Known as the Peruvian Coastal Upwelling
Region (PCUR), the waters east of the Galápagos are
among the coldest and most productive within the ETP
(Pennington et al. 2006). Considering that this area
was dominated by waters ≤25°C during the 2005 tracking season (Fig. 6), this temperature may represent the
thermal threshold below which migrating green turtles
actively avoid surface waters. We acknowledge that
the lack of movements into the PCUR may be an artifact of small sample size or perhaps because accessing
this area would require swimming against the prevailing SEC and Peru Current. However, we doubt that
surface currents were a major factor in this lack of eastward movements since green turtles were found to
swim countercurrent in other areas during this study.
With respect to SST, Godley et al. (2002) reported a
marked shift in diving behavior by migrating green
turtles in the Mediterranean when mean sea temperatures dropped below 25°C, further supporting the likelihood that green turtles from the Galápagos avoided
the PCUR due to its low surface temperatures. Interestingly, turtles in the Galápagos neritic appear willing to
experience colder water, perhaps the result of a
greater thermal tolerance during internesting behavior, but also possibly due to smaller-scale movements
by turtles in these areas and the correspondingly lesser
opportunity to choose specific water temperatures. It
would be interesting to conduct telemetry studies of
Galápagos green turtles during periods of elevated
SST (i.e. ENSO) to determine if turtles are more likely
to access waters east of the Galápagos during these
periods. If so, these warming episodes may provide a
mechanism for green turtles to access foraging habitats
along the coast of South America.
In addition to thermal conditions, sea surface chl a
concentration appears to be an oceanographic variable
that influenced green turtle movements. Oceanic
migrations occurred in waters with a mean sea surface
chl a concentration of 0.18 mg m– 3 and Argos locations
were tightly grouped around the 0.2 mg
m– 3 surface chl a contour. This is consistent with satellite-tracked movements
of loggerhead turtles foraging along the
North Pacific transition zone chlorophyll front (TZCF) (Polovina et al. 2001).
The 0.2 mg m– 3 surface chl a contour is
considered a proxy for identifying the
convergence zone between the stratified low surface chl a (< 0.15 mg m– 3)
waters and the high surface chl a
(> 0.3 mg m– 3) vertically mixed waters
(Polovina et al. 2000, 2001). This convergence zone is biologically important
since the intersection of these 2 chlorophyll masses often results in the increased abundance of zooplankton and
other buoyant organisms (Olson et al.
1994, Polovina et al. 2000, 2001, Graham et al. 2001) that constitute important food resources for marine predators. However, it should also be noted
that the attenuation of light underwater
increases exponentially with surface chl
a (Smith & Baker 1978), the result being
that waters with a surface chl a concentration of 0.2 mg m– 3 are relatively clear,
and likely beneficial to visual predators
such as sea turtles. Thus, the affinity of
Fig. 6. Chelonia mydas. Green turtle post-nesting paths (continuous black lines)
in 2005 in relation to frontal zones and SST contour locations (dashed lines)
green turtles to waters with a 0.2 mg
Seminoff et al.: Post-nesting migrations of Chelonia mydas
13
m– 3 surface chl a concentration may
simply indicate their preference for
clear water.
The localized movements of green turtles in areas south of the Galápagos, coupled with their association with preyconducive surface chlorophyll levels
and SST fronts, underscores the likelihood that at least some green turtles foraged in oceanic waters during tracking
Fig. 7. Jurisdictional waters occupied by green turtles during post-nesting
efforts. Oceanic foraging by green turmovements tracked with satellite telemetry
tles has been demonstrated via stable
isotope analyses by Hatase et al. (2006).
Rican longline fisheries operating east of the GalápaParker & Balazs (in press) analyzed stomachs of green
gos. Green turtles are also the second most common
turtles bycaught in oceanic zones of the central Pacific
sea turtle species bycaught in Peruvian artisanal shark
and found a variety of oceanic prey species such as sea
longline fisheries (Kelez-Sara et al. 2006) and Colomjellies Pyrosoma sp., commonly found at the surface,
bian shrimp trawl fisheries (Amorocho et al. 2005), and
gooseneck barnacles Lepas sp., which adhere to floathave been killed as the result of bycatch in artisanal
ing objects, and Janthina sp., a gastropod that occurs at
fisheries in Panama (P. Zarate unpubl. data) and Chile
the sea surface with the assistance of a gas-filled float,
(M. Donoso pers. comm.). Together, these threats may
as well as > 20 additional oceanic species. Green turtles
have substantial negative effects on the Galápagos
captured in pelagic zones off mainland Ecuador and
nesting population. A decline in the annual number of
Peru have had stomachs filled with fish eggs (Fritts
nesting females from 2000 to 2006 at the 4 primary
1981b, Hays-Brown & Brown 1982). However, despite
beaches (Zárate 2007) suggests that the consequences
this evidence of oceanic foraging by green turtles, the
of these effects are apparent at the nesting beaches.
strategies used to access oceanic prey resources remain
Global conservation efforts have largely focused on
poorly understood. To study this behavior we encournesting beaches, representing only a fraction of marine
age additional satellite telemetry efforts that employ
turtle life history, but it is now clear that safeguarding
GPS and dive-profiling technology to reveal the exact
the migratory corridors and neritic foraging areas is
locations of green turtles in relation to oceanographic
essential. Fortunately, there are several sea turtle confeatures as well as depict the actual diving activity and
servation initiatives ongoing in the ETP that may help
surfacing intervals in these areas.
protect sea turtles. Those turtles traveling north from
the Galápagos to areas near Cocos Island (Costa Rica),
Coiba Island (Panama), and Gorgona and Malpelo
Conservation implications
Islands (Colombia) are legally protected by the Marine
Conservation and Sustainable Development Corridor,
Green turtles nesting within the Galápagos are proan initiative of the governments of Costa Rica, Panama,
tected by the Galápagos Biological Reserve of Marine
Colombia, and Ecuador which promotes sustainable
Resources; however, after nesting, the majority of turmanagement of marine resources within this part of
tles swim to distant foraging areas that have less highly
the ETP. The Eastern Tropical Pacific Seascape Project
regulated and largely unmanaged exploitation of sea
is a complementary effort by non-governmental orgaturtles (e.g. Chacón 2002). Turtles tracked during the
nizations which aims to protect a variety of marine verpresent study traveled for substantial periods in intertebrate species within this same area (UNESCO 2006).
national waters and entered jurisdictional waters of 5
Additional international instruments such as the InterLatin American nations (Fig. 7), all of which have been
American Tropical Tuna Commission (IATTC) and
shown to have ongoing human impacts on sea turtles
Inter-American Convention for the Protection and
in neritic foraging areas (Cornelius 1982, Chacón 2002,
Conservation of Sea Turtles (IAC) are also designed to
Amorocho et al. 2005). Although not visited by green
lessen the effects on sea turtles from fisheries and
turtles during this study, coastal regions in Peru and
other human impacts. While clearly no single law or
mainland Ecuador also have high levels of directed
treaty can be 100% effective at minimizing anthrotake (de Paz et al. 2002, M. Helvey pers. comm.), furpogenic impacts on sea turtles, these international
ther underscoring the widespread nature of green turagreements and laws provide a framework within
tle hunting in the ETP. With regard to bycapture in
which conservation advances can be made. However,
fisheries gear, Arauz et al. (2000) describe a catch per
there is also a need for a broad-based, cohesive sea
unit effort of 0.85 green turtles per 100 hooks in Costa
14
Endang Species Res 3: preprint, 2007
turtle conservation plan that targets all sea turtle species occurring in the ETP. The Permanent Commission
of the South Pacific, under the Lima Convention, has
developed an Action Plan for Sea Turtles in the Southeast Pacific (CPPS 2006), and is perhaps the most relevant of the international instruments in this region
since all 5 nations from Panama to Chile are signatories. However, the successful implementation of this
plan will require additional data on sea turtle biology
and fisheries impacts as well as explicit discussion of
the viable solutions to the ongoing problems. Clearly,
such an effort would benefit from the participation and
input of a diverse group including sea turtle biologists,
wildlife managers, economists, fishers, and policy
experts to address the growing challenges.
Acknowledgements. We thank C. Chasiluisa, M. Chica,
M. Farias, C. Hasbun, R. LeRoux, and N. de Paz for their assistance with this project. Logistic and financial support was
provided by the Charles Darwin Research Station and
NOAA-National Marine Fisheries Service. We also thank
R. Arauz, G. Balazs, M. Chaloupka, B. Godley, R. Kennett,
and Y. Matsuzawa for providing information for this paper.
B. Godley, B. Wallace and 4 anonymous reviewers provided
valuable comments that helped improve earlier versions of
this manuscript. Research permits were provided by the Galápagos Biological Reserve of Marine Resources.
Dedication. This article is dedicated to our friend and colleague Boyd Nathaniel Lyon (1969–2006). Boyd was taken
from us much too early, while doing one of the things he loved
most: researching the in-water ecology of green turtles. He
was among the best and brightest rising stars and was able to
touch the lives of many fellow researchers during his years as
part of our research community. His passion, attention to
detail, and robust spirit is an inspiration to all.
LITERATURE CITED
Alvarado J, Figueroa A (1992) Recapturas post-anidatorias de
hembras de tortuga marina negra (Chelonia agassizi) marcadas en Michoacan, Mexico. Biotropica 24:560–566
Amorocho DF, Reina RD (2007) Feeding ecology of the East
Pacific green sea turtle Chelonia mydas agassizii at Gorgona National Park, Colombia. Endang Species Res 3:43–51
Amorocho D, Barreto L, Zapata L (2005) Disminución del impacto por captura incidental en tortugas marinas con el uso
de anzuelos curvos.Informe Técnico WWF-Colombia, Cali
Arauz R, Rodríguez O, Vargas R, Segura A (2000) Incidental
capture of sea turtles by Costa Rica’s longline fleet. In:
Kalb H, Wibbels T (compilers) Proceedings of the 19th
Annual Symposium on Sea Turtle Biology and Conservation. NOAA Tech Memo NMFS-SEFSC-443, National
Marine Fisheries Service, Miami, FL, p 21–26
Balazs GH, Miya RK, Beaver SC (1996) Procedures to attach a
satellite transmitter to the carapace of an adult green sea
turtle, Chelonia mydas. In: Keinath JA, Barnard DE,
Musick JA, Bell BA (compilers) Proceedings of the 15th
Annual Symposium on Sea Turtle Biology and Conservation, NOAA Tech Memo, NMFS-SEFSC-387, National
Marine Fisheries Service, Miami, FL, p 21–26
Bass AL, Epperly SP, Braun-McNeill J (2006) Green turtle
(Chelonia mydas) foraging and nesting aggregations in
the Caribbean and Atlantic: impact of currents and behavior on dispersal. J Hered 97:346–354
Bjorndal KA (1985) Nutritional ecology of sea turtles. Copeia
1985:736–751
Bjorndal KA, Bolten AB, Chaloupka MY (2000) Green turtle
somatic growth model: evidence for density dependence.
Ecol Appl 10:269–282
Blumenthal JM, Solomon JL, Bell CD, Austin TJ and others
(2006) Satellite tracking highlights the need for international cooperation in marine turtle management. Endang
Species Res 2:51–61
Breaker LC, Mavor TP, Broenkow WW (2005) Mapping and
monitoring large-scale ocean fronts off the California
Coast using imagery from the GOES-10 geostationary
satellite. Publ T-056, California Sea Grant College Program, University of California, San Diego, La Jolla, CA
Broderick AC, Coyne MS, Fuller WJ, Glen F, Godley BJ
(2007) Fidelity and over-wintering of sea turtles. Proc R
Soc Lond B Biol Sci 274:1533–1538
Canny J (1986) A computational approach to edge-detection.
IEEE Trans Pattern Anal Mach Intell 8:679–698
Carr A, Meylan AB (1980) Evidence of passive migration of
green turtle hatchlings in Sargassum. Copeia 1980: 366–368
Carr A, Carr MH, Meylan AB (1978) The ecology and migrations of sea turtles, 7. The west Caribbean green turtle
colony. Bull Am Mus Nat Hist 162:1–46
Castelao RM, Barth JA, Mavor TP (2005) Flow-topography
interactions in the northern California Current System
observed from geostationary satellite data. Geophys Res
Lett 32:L24612
Cayula JF, Cornillon P (1995) Multi-image edge detection for
SST images. J Atmos Ocean Technol 12:821–829
Chacón D (2002) Assessment about the trade of sea turtles
and their products in the Central American isthmus. Red
Regional para la Conservación de last Tortugas Marinas
en Centroamérica. San José, Costa Rica
Chavez FP, Strutton PG, Friederich GE, Feely RA, Feldman
GC, Foley DG, McPhaden MJ (1999) Biological and chemical response of the equatorial Pacific Ocean to the
1997–98 El Niño. Science 286:2126–2131
Cheng IJ (2000) Post nesting migrations of green turtles (Chelonian mydas) at Wan-An Island, Penghu Archipelago,
Taiwan. Mar Biol 137:747–754
Cornelius SE (1982) Status of sea turtles along the Pacific
coast of Middle America. In: Bjorndal KA (ed) Biology and
conservation of sea turtles. Smithsonian Institute Press,
Washington, DC, p 211–219
Coyne MC, Godley BJ (2005) Satellite Tracking and Analysis
Tool (STAT): an integrated system for archiving, analyzing
and mapping animal tracking data. Mar Ecol Prog Ser 301:
1–7
CPPS (Comisión Permanente del Pacífico Sur) (2006) Plan de
Acción para la Protección del Medio Marino y Áreas
Costeras del Pacífico Sudeste. Comisión Permanente del
Pacífico Sur, Guayaquil, Ecuador
Craig P, Parker D, Brainard R, Rice M, Balazs G (2004)
Migrations of green turtles in the central South Pacific.
Biol Conserv 116:433–438
de Paz N, Reyes J, Echegaray M (2002) Datos sobre captura,
comercio y biología de tortugas marinas en el área de
Pisco-Paracas. In: Mendo J, Wolf M (eds) Jornada Científica Reserva Nacional de Paracas. Universidad Nacional
Agraria La Molina, Lima, p 125–129
Dethmers KEM, Broderick D, Moritz C, FitzSimmons NN and
others (2006) The genetic structure of Australasian green
turtles (Chelonia mydas): exploring the geographical scale
of genetic exchange. Mol Ecol 15:3931–3946
Seminoff et al.: Post-nesting migrations of Chelonia mydas
Doniol-Valcroze T, Berteaux D, Larouche P, Sears R (2007)
Influence of thermal fronts on habitat selection by four
rorqual whale species in the Gulf of St. Lawrence. Mar Ecol
Prog Ser 335:207–216
Ducet N, LeTraon PY, Reverdin G (2000) Global high resolution mapping of ocean circulation from Topex/Poseidon
and ERS-1 and -2. J Geophys Res 105:19477–19498
Eden C, Timmerman A (2004) The influence of the Galápagos
Islands on tropical temperatures, currents, and the generation of tropical instability waves. Geophys Res Lett 31:
L15308
Etnoyer P, Canny D, Mate BR, Morgan LE, Ortega-Ortiz JG,
Nichols WJ (2006) Sea-surface temperature gradients
across blue whale and sea turtle foraging trajectories off
the Baja California Peninsula, Mexico. Deep-Sea Res II
53:340–358
Fiedler PC (2002a) Environmental change in the eastern
tropical Pacific Ocean: review of ENSO and decadal
variability. Mar Ecol Prog Ser 244:265–283
Fiedler PC (2002b) The annual cycle and biological effects of
the Costa Rica Dome. Deep-Sea Res I 49:321–338
Fiedler PC, Bernard HJ (1987) Tuna aggregation and feeding
near fronts observed in satellite imagery. Cont Shelf Res 7:
871–881
Fritts TH (1981a) Marine turtles of the Galápagos Islands
and adjacent areas of the eastern Pacific on the basis of
observations made by J. R. Slevin 1905–1906. J Herpetol
15:293–301
Fritts TH (1981b) Pelagic feeding habits of turtles in the eastern Pacific. Mar Turt Newsl 17:4–5
Godley BJ, Richardson S, Broderick AC, Coyne MS, Glen F,
Hays GC (2002) Long-term satellite telemetry of the
movements and habitat utilization by green turtles in the
Mediterranean. Ecography 25:352–362
Godley BJ, Broderick AC, Glen F, Hays GC (2003) Post-nesting movements and submergence patterns of loggerhead
marine turtles in the Mediterranean assessed by satellite
tracking. J Exp Mar Biol Ecol 287:119–134
Godley BJ, Blumenthal JM, Broderick AC, Coyne MS, Godfrey MH, Hawkes LA, Witt MJ (2007) Satellite tracking of
sea turtles: Where have we been and where do we go
next? Endang Species Res N 60 (in press)
Graham WM, Pagès F, Hamner WM (2001) A physical context
for gelatinous zooplankton aggregations: a review. Hydrobiologia 451:199–212
Green DJ (1984) Long-distance movements of Galápagos
green turtles. J Herpetol 18:121–130
Green DJ (1993) Growth rates of wild immature green turtles
in the Galápagos Islands, Ecuador. J Herpetol 27:338–341
Green D, Ortiz-Crespo F (1982) Status of sea turtle populations in the central eastern Pacific. In: Bjorndal KA (ed),
Biology and conservation of sea turtles. Smithsonian Institution Press, Washington, DC, p 221–234
Hatase H, Takai N, Matsuzawa Y, Sakamoto W, Omuta K,
Goto K, Arai K, Fujiwara T (2002) Size-related differences
in feeding habitat use of adult female loggerhead turtles
Caretta caretta around Japan determined by stable isotope analyses and satellite telemetry. Mar Ecol Prog Ser
233:273–281
Hatase H, Sato K, Yamaguchi M, Takahashi K, Tsukamoto K
(2006) Individual variation in feeding habitat use by adult
green sea turtles (Chelonia mydas): are they obligately
neritic herbivores? Oecologia 149:52–64
Hawkes LA, Broderick AC, Coyne MS, Godfrey MS and others (2006) Phenotypically linked dichotomy in sea turtle
foraging requires multiple conservation approaches. Curr
Biol 16:990–995
15
Hawkes LA, Broderick AC, Coyne MS, Godfrey MH, Godley
BJ (2007) Only some like it hot — quantifying the environmental niche of the loggerhead sea turtle. Divers Distrib 13:
447–457
Hays GC, Dray M, Quaife T, Smyth TJ and others (2001) Movements of migrating green turtles in relation to AVHRR derived
sea surface temperature. Int J Remote Sens 22: 1403–1411
Hays GC, Luschi P, Papi F, del Seppia C, Marsh R (2002)
Bi-phasal long-distance migration in green turtles. Anim
Behav 64:895–898
Hays-Brown C, Brown WM (1982) Status of sea turtles in the
southeastern Pacific: emphasis on Peru. In: Bjorndal KA
(ed) Biology and conservation of sea turtles. Smithsonian
Institution Press, Washington, DC, p 235–240
Heylings P, Bensted-Smith R, Altamirano M (2002) Zonificación e historia de la Reserva Marina de Galápagos. In:
Danulat E, Edgar GJ (eds) Reserva Marina de Galápagos,
Línea Base de la Biodiversidad. Fundación Charles Darwin/Servicio Parque Nacional Galápagos, Santa Cruz,
Galápagos, p 10–37
Hollander M, Wolfe DA (1999) Nonparametric statistical
methods, 2nd edn. Wiley & Sons, New York
IOC IHO, BODC (2003) Centenary edition of the GEBCO digital atlas. British Oceanographic Data Centre, Liverpool
Kelez-Sara S, Manrique-Bravo C, Velez-Zuazo X (2006)
Shark longline fishey and sea turtles in Peruvian waters.
In: Frick M, Panagopoulou A, Rees AF, Williams K (compilers). Book of Abstracts; 26th Annual Symposium on Sea
Turtle Biology and Conservation. International Sea Turtle
Society, Athens, p 262–263
Kennett R, Munungurritj N, Yunupingu D (2004) Migration
patterns of marine turtles in the Gulf of Carpinteria, northern Australia: implications for aboriginal management.
Wildl Res 31:241–248
Kessler WS (2006) The circulation of the eastern tropical
Pacific: a review. Prog Oceanogr 69:181–217
Kuroyanagi K, Kamezaki N, Sugiyama S (1999) First confirmed black sea turtle (C. m. agassizii) in Japanese waters.
Umigame Newsl Jpn 39:13–14
Limpus CJ, Miller J, Parmenter CJ, Reimer D, McLachlan N,
Webb R (1992) Migration of green (Chelonia mydas) and
loggerhead (Caretta caretta) turtles to and from eastern
Australian rookeries. Wildl Res 19:347–358
Luschi P, Hays GC, del Seppia C, Marsh R, Papi F (1998) The
navigational feats of green sea turtles migrating from
Ascension Island investigated by satellite telemetry. Proc
R Soc Lond B Biol Sci 265:2279–2284
Luschi P, Hays GC, Papi F (2003) A review of long-distance
movements by marine turtles, and the possible role of
ocean currents. Oikos 103:293–302
Morreale SJ, Standora EA, Spotila JR, Paladino FV (1996)
Migration corridor for sea turtles. Nature 384:319–320
Musick JA, Limpus CJ (1997) Habitat utilization and migration
in juvenile sea turtles. In: Lutz PL, Musick JA (eds) The biology of sea turtles. CRC Press, Boca Raton, FL, p 137–163
Olson DB, Hitchcock GL, Mariano AJ, Ashjian CJ, Peng G,
Nero RW, Podesta GP (1994) Life on the edge: marine life
and fronts. Oceanography 7:52–59
Palacios DM (2002) Factors influencing the island-mass effect
of the Galápagos Islands. Geophys Res Lett 29:2134
Palacios DM, Bograd SJ, Foley DG, Schwing FB (2006)
Oceanographic characteristics of biological hot spots in
the North Pacific: a remote sensing perspective. Deep-Sea
Res II 53:250–269
Parker DM, Balazs GH (in press) Diet of the oceanic green turtle, Chelonia mydas, in the North Pacific. In: Proceedings
of the 25th International Symposium on Sea Turtle Biology
16
Endang Species Res 3: preprint, 2007
and Conservation, National Marine Fisheries Service,
Miami, FL
Pelletier D, Roos D, Ciccione S (2003) Oceanic survival and
movements of wild and captive reared immature green turtles (Chelonia mydas) in the Indian Ocean. Aquat Living
Resour 16:35–41
Pennington JT, Mahoney KL, Kuwahara VS, Kolber DD,
Calienes R, Chavez FP (2006) Primary production in the
eastern tropical Pacific: a review. Prog Oceanogr 69:
285–317
Plotkin P (2003) Adult migrations and habitat use. In: Lutz PL,
Musick JA, Wyneken J (eds) The biology of sea turtles,
Vol 2. CRC Press, Boca Raton, FL, p 225–242
Plotkin PT, Byles RA, Rostal DC, Owens DW (1995) Independent versus socially facilitated oceanic migrations of the
olive ridley, Lepidochelys olivacea. Mar Biol 122:137–143
Podesta GP, Brower JA, Hoey JJ (1993) Exploring the association between swordfish catch rates and thermal fronts on
United States longline grounds in the Western NorthAtlantic. Cont Shelf Res 13:253–277
Polovina JR, Kobayashi DR, Parker DM, Seki JP, Balazs GH
(2000) Turtles on the edge: movements of loggerhead turtles (Caretta caretta) along oceanic fronts, spanning longline fishing grounds in the central North Pacific,
1997–1998. Fish Oceanogr 9:71–82
Polovina JJ, Howell EE, Kobayashi DR, Seki MP (2001) The
transition zone chlorophyll front, a dynamic global feature
defining migration and forage habitat for marine
resources. Prog Oceanogr 49:469–483
Polovina J, Uchida I, Balazs G, Howell EA, Parker D, Dutton
P (2006) The Kuroshio Extension Bifurcation Region: a
pelagic hotspot for juvenile loggerhead sea turtles.
Deep-Sea Res II 53:326–339
Saba VS, Santidrián-Tomillo P, Reina RD, Spotila JR, Musick
JA, Evans DA, Paladino FV (2007) The effect of the El
Niño Southern Oscillation on the reproductive frequency
of eastern Pacific leatherback turtles. J Appl Ecol 44:
395–404
Smith RC, Baker KS (1978) The bio-optical state of ocean
waters and remote sensing. Limnol Oceanogr 23:
247–259
Strub PT, Mesias J, James C (1995) Satellite observations of the
Peru–Chile countercurrent. Geophys Res Lett 22:211–214
Strub PT, Mesias JM, Montecino V, Rutllant J, Salinas S
(1998) Coastal ocean circulation off Western South America. In: Robinson AR, Brink KH (eds) The sea (coastal
oceans), Vol 11. Wiley, New York, p 273–313
Troëng S, Evans DR, Harrison E, Lagueux CJ (2005) Migration of green turtles Chelonia mydas from Tortuguero,
Costa Rica. Mar Biol 148:435–447
UNESCO (2006) Eastern Tropical Pacific Seascape Project.
Available at: http://whc.unesco.org/en/activities/14/
(Downloaded on 30 October 2006)
Wessel P, Smith WHF (1991) Free software helps map and display data. Eos 72:445–446
Whiting SD, Long JL, Coyne M (2007) Migration routes
and foraging behaviour of olive ridley turtles Lepidochelys
olivacea in northern Australia. Endang Species Res 3:1–9
Worm B, Sandow M, Oschlies A, Lotze HK, Myers RA (2005)
Global patterns of predator diversity in the open oceans.
Science 309:1365–1369
Zárate P (2007) Assessment of the foraging areas of marine
turtles in the Galápagos Islands: 2000–2006. Final report
to NOAA-National Marine Fisheries Service. Charles Darwin Foundation, Puerto Ayora, Galápagos
Zárate P, Dutton P (2002) Tortuga verde. In: Danulat E, Edgar
GJ (eds) Reserva Marina de Galápagos. Línea base de
la biodiversidad. Fundación Charles Darwin/Servicio
Parque Nacional Galápagos, Santa Cruz, Galápagos,
p 305–323
Appendix 1. Satellite transmitter performance for 12 deployments on green turtles Chelonia mydas. Data for each individual includes date of deployment, start and end dates of satellite transmissions, days at large, and number of satellite uplinks summarized
by location class (LC). BA: Barahona; LS: Las Salinas; QP: Quinta Playa; see Fig. 1 for tagging site locations
Turtle Tagging
site
CM-1
CM-2
CM-3
CM-4
CM-5
CM-6
CM-7
CM-8
CM-9
CM-10
CM-11
CM-12
BA
BA
LS
LS
LS
QP
QP
QP
QP
QP
QP
BA
Date of
release
Date of
first location
25 Feb 03
27 Feb 03
2 Mar 03
2 Mar 03
1 Apr 05
8 Apr 05
9 Apr 05
9 Apr 05
9 Apr 05
10 Apr 05
10 Apr 05
11 Apr 05
28 Feb 03
1 Mar 03
9 Mar 03
2 Mar 03
2 Apr 05
10 Apr 05
9 Apr 05
10 Apr 05
10 Apr 05
11 Apr 05
11 Apr 05
11 Apr 05
Date of
Track
final location duration (d)
17 Apr 03
27 Apr 03
10 Jun 03
29 Apr 03
2 Jun 05
11 May 05
10 Jun 05
19 Jun 05
18 Jun 05
15 Jun 05
20 Jun 05
18 Jul 05
Total
Editorial responsibility: Brendan Godley (Editor-in-Chief),
University of Exeter, Cornwall Campus, UK
3
2
1
LC
0
A
B
Z
Total
uplinks
51
59
100
61
62
33
62
69
70
65
67
108
0
0
1
2
1
0
2
1
2
3
0
3
1
3
2
6
8
1
3
2
8
6
2
6
4
5
11
29
9
9
5
5
10
5
4
11
6
10
9
26
14
9
8
4
20
2
4
7
9
19
18
44
12
6
7
13
21
15
8
36
16
29
46
67
30
14
18
22
32
26
15
55
1
3
5
10
2
5
1
1
4
0
1
4
37
69
92
184
76
44
44
48
97
57
34
122
807
15
48
107
119
208
370
37
904
Submitted: June 26, 2007; Accepted: October 31, 2007
Proofs received from author(s): November 27, 2007